I highlight three results from cosmological hydrodynamic simulations that yield a realistic red sequence of galaxies: 1) Major galaxy mergers are not responsible for shutting off star-formation and forming the red sequence. Starvation in hot halos is. 2) Massive galaxies grow substantially (~ × 2 in mass) after being quenched, primarily via minor (1:5) mergers. 3) Hot halo quenching naturally explains why galaxies are red when they either (a) are massive or (b) live in dense environments.

We demonstrate that massive simulated galaxies assemble in two phases, with the initial growth dominated by compact in situ star formation, whereas the late growth is dominated by accretion of old stars formed in subunits outside the main galaxy. We also show that 1) gravitational feedback strongly suppresses late star formation in massive galaxies contributing to the observed galaxy colour bimodality that 2) the observed galaxy downsizing can be explained naturally in the two-phased model and finally that 3) the details of the assembly histories of massive galaxies are directly connected to their observed kinematic properties.

According to a popular scenario supported by numerical models, the mass assembly and growth of massive galaxies, in particular the Early-Type Galaxies (ETGs), is, below a redshift of 1, mainly due to the accretion of multiple gas–poor satellites. In order to get observational evidence of the role played by minor dry mergers, we are obtaining extremely deep optical images of a complete volume limited sample of nearby ETGs. These observations, done with the CFHT as part of the ATLAS3D, NGVS and MATLAS projects, reach a stunning 28.5 – 29 mag.arcsec−2 surface brightness limit in the g' band. They allow us to detect the relics of past collisions such as faint stellar tidal tails as well as the very extended stellar halos which keep the memory of the last episodes of galactic accretion. Images and preliminary results from this on-going survey are presented, in particular a possible correlation between the fine structure index (which parametrizes the amount of tidal perturbation) of the ETGs, their stellar mass, effective radius and gas content.

The formation of the brightest cluster galaxies (BCG) is a challenge for galaxy formation theory. We performed high resolution cosmological hydrodynamical simulations with the AMR code RAMSES to study the properties of the BCG which forms at the center of a Virgo–like cluster. We compare the results of 2 galaxy formation scenarios, one in which only supernovae feedback is included, and one in which also AGN feedback is considered. Properties of the simulated BCG which are comparable with those of observed massive elliptical galaxies and BCGs cannot be obtained if AGN feedback is not considered. The stellar-to-halo mass ratio in simulations without AGN feedback appears too large when compared to observations, while it is compatible the observationally determined values when AGN feedback is included. The kinematical and structural properties of the BCG are extremely different in the two models. When we do not include AGN feedback, the BCG is quickly rotating, with high Sérsic index, a clear mass excess in the center and a very large stellar mass fraction. When AGN feedback is considered, the BCG is slowly rotating, with a significantly cored surface density profile and low stellar mass fraction.

In this contribution we present the results of an Eulerian adaptive mesh refinement (AMR) hydrodynamical and N-body simulation in a ΛCDM cosmology. The simulation used was performed with the cosmological code MASCLET (Quilis et al. 2004). Galaxies have been identified in the simulation outputs by means of an adaptive friends of friends algorithm applied to the star particles. To give light to our virtual galaxies we have assigned a spectrum to each stellar particle using the MIUSCAT stellar population models (Vazdekis et al. 2012; Ricciardelli et al. 2012).

Future prospects in observational galaxy evolution are reviewed from a personal perspective. New insights will especially come from high-redshift integral field kinematic data and similar low-redshift observations in very large and definitive surveys. We will start to systematically probe the mass structures of galaxies and their haloes via lensing from new imaging surveys and upcoming near-IR spectroscopic surveys will finally obtain large numbers of rest frame optical spectra at high-redshift routinely. ALMA will be an important new ingredient, spatially resolving the molecular gas fuelling the high star-formation rates seen in the early Universe.

This is a brief introduction to the closing discussion of the IAU Symposium 295, ‘The Intriguing Life of Massive Galaxies’, that was held in Beijing from August 27 through 31, 2012. The discussion was focused on only four hot items, namely 1) the redshift evolution of the size of passively evolving galaxies, 2) the evolution with redshift of the specific star formation rate, 3) quenching of star formation in galaxies and dry merging, and 4) the IMF.